Solving the Puzzle of Cognitive Reserve Effects on Cognitive Decline: The Importance of Considering Functional Impairment

In the context of cognitive impairment such as dementia, cognitive reserve is a crucial factor [1-3]. In general, the cognitive reserve concept [4-6] aims to account for individual differences in the course of aging and neurodegenerative development. Cognitive reserve is built up during life through cognitive stimulation and serves later on as buffer for coping with brain alterations in order to preserve cognitive functioning in aging [4-6]. Frequently used proxies of accumulated cognitive reserve across the lifespan are education, cognitively demanding jobs, and leisure activity engagement [4-7]. Empirically, cross-sectional evidence demonstrated that these cognitive-reserve proxies are correlated with better cognitive functioning level in healthy older adults [7-9].

Yet, empirical findings on the longitudinal relation of cognitive reserve to long-term rate of cognitive decline are highly inconsistent and, therefore, remain inconclusive so far. For instance, some studies found relations of cognitive reserve markers (e.g., higher education and greater leisure activity engagement) to reduced cognitive decline [10-12], while others observed the opposite pattern, that is, steeper cognitive decline in individuals with greater cognitive reserve [13-16]. Yet, other studies found relations only to better cognitive performance level, but not to cognitive decline [17-20] (see [21-24] for discussions).

One avenue in solving this highly debated puzzle of inconsistent results concerns the question whether the relation between cognitive reserve and the rate of cognitive decline may be moderated by certain individual-difference characteristics. For example, in the context of severe cognitive decline and cognitive impairment, consequences with regard to functional impairment, such as difficulties in independently performing one’s activities of daily living, are a crucial issue [25-27]. Interestingly, recent cross-sectional evidence suggests that correlations of cognitive-reserve markers and cognitive functioning level may be stronger in individuals with poorer, compared to those with better functional fitness status (such as strength, agility, and endurance) [28, 29]. However, so far research was only cross-sectional and considered rather momentary markers of cognitive reserve. To advance our understanding in this regard, a longitudinal investigation on long-term cognitive decline comprising the cumulative nature of cognitive reserve build-up during life is essential. Therefore, we investigated whether the longitudinal relation between cognitive reserve accumulated across the lifespan and rate of cognitive decline over 6 years as measured through performance changes in the Trail Making Test (TMT) differed by the individual’s degree of functional impairment.

We analyzed data from 897 individuals who participated in the 2 waves of the Vivre-Leben-Vivere (VLV) survey [30-32]. Respondents were first interviewed during 2011 (Wave 1; W1) and again in 2017 (Wave 2; W2) using face-to-face computer-assisted personal interviewing (CAPI) and paper-pencil questionnaires. For further details regarding the rationale, design, recruitment, materials, and procedures of the VLV survey see [30-39]. Mean age of these respondents in W1 was 74.33 years (SD = 6.50, range 64–96); 51.4% were men.

In both waves, we administered the TMT parts A and B (TMT A and TMT B, respectively [40]). Functional impairment in W1 was indicated by difficulties in independently managing one’s activities of daily living, such as dressing, feeding, washing, and transferring [41, 42]. We used the Cognitive Reserve Index questionnaire [43] to assess proxies of accumulated cognitive reserve during life (comprising education, cognitive demand of jobs, and leisure activity engagement). To obtain an overall indicator of cognitive reserve accumulated across the lifespan, we computed a total cognitive reserve score (see [43, 44] for a detailed description). We controlled for the overall number of chronic diseases (such as heart diseases of ischemic or organic pathogenesis, primary arrhythmias, pulmonary heart diseases, hypertension, and peripheral vascular diseases) participants suffered from in W1 [45].

Using latent change score modeling, we modeled latent cognitive factors of TMT completion time in W1 (constructed from TMT parts A and B in W1) and W2 (constructed from TMT parts A and B in W2) as well as a latent cognitive change variable regarding change in TMT completion time from W1 to W2 [45, 46]. We included the following covariates to predict latent change: cognitive reserve, functional impairment in W1, the number of chronic diseases in W1, age in W1, sex, and the interaction of functional impairment in W1 with cognitive reserve (while taking the dependencies among all covariates into account). Data are available online as supplemental material (see online suppl. Table 1; see www.karger.com/doi/10.1159/000511768 for all online suppl. material). For model estimation, we used full information maximum likelihood.

Greater functional impairment in W1 (β = 0.20, p < 0.001), a larger number of chronic diseases in W1 (β = 0.10, p = 0.012), and older age in W1 (β = 0.30, p < 0.001) significantly predicted a larger increase in TMT completion time from W1 to W2 (i.e., steeper cognitive decline). Sex and cognitive reserve total score per se did not predict changes in TMT completion time (p > 0.05). Yet, most importantly, there was a significant interaction of functional impairment with cognitive reserve total score (β = 0.18, p < 0.001).

We additionally explored this interaction pattern and disentangled the specific contributions by the 3 cognitive reserve markers. There was no interaction of functional impairment with education (β = 0.00, p = 0.917). Notably, there were significant interactions of functional impairment with cognitive demand of jobs (β = 0.18, p < 0.001) and of functional impairment with leisure activity engagement (β = 0.15, p < 0.001).

To describe these interactions, we estimated in our latent change score model the longitudinal relation between cognitive reserve and changes in TMT completion time at different degrees of functional impairment. Specifically, at an impairment score of zero in W1 (i.e., no functional impairment), greater cognitive reserve accumulated through leisure engagement significantly predicted a smaller increase in TMT completion time from W1 to W2 (i.e., reduced cognitive decline, β = −0.15, p < 0.001). In contrast, at an impairment score of at least 2 points in W1 (i.e., incapability, i.e., entire dependence in at least 1 activity; or moderate impairment in at least 2 activities such as dressing, feeding, washing, or transferring [42]), greater cognitive reserve accumulated through leisure engagement significantly predicted a larger increase in TMT completion time from W1 to W2 (i.e., steeper cognitive decline, β ≥0.30, p < 0.027). Notably, already at an impairment score of at least 1 point in W1 (i.e., moderate impairment in at least 1 activity [42]), greater cognitive reserve accumulated through cognitive demand of jobs significantly predicted a larger increase in TMT completion time from W1 to W2 (i.e., steeper cognitive decline, β ≥0.26, p < 0.001).

With regard to the highly inconsistent and debated puzzle of results observed in prior research on the relation between cognitive reserve and rate of cognitive decline, present longitudinal results have important implications. Using latent change score modeling (extracting measurement-error variance), we demonstrated that the individual’s degree of functional impairment substantially moderates the longitudinal relation between cognitive-reserve built-up during life and cognitive decline over 6 years. Specifically, with no functional impairment in the first wave of assessment, greater cognitive reserve accumulated across the lifespan significantly predicted a reduced cognitive decline over 6 years (i.e., indicated by a smaller increase in TMT completion time). In contrast, with certain functional impairment (in at least some activities) in the first wave, greater cognitive-reserve build-up predicted a steeper cognitive decline (i.e., indicated by a larger increase in TMT completion time).

On first glance perhaps, the latter finding may sound counterintuitive. Yet, the overall pattern observed actually dovetails with the theoretical predictions of the cognitive reserve concept [4, 5]: individuals with greater cognitive reserve accumulated are able to tolerate more pathology, that is, they can still maintain cognitive functioning for a longer time and thereby show a reduced cognitive decline (though pathology is advancing). Cognitive functioning will begin to decline relatively late in time, after more pathology has accumulated. But, once decline begins, the cognitive system rapidly collapses and will then decline at a much steeper rate given that pathology has already been highly advanced [4, 5].

Importantly, our longitudinal observations empirically corroborate these conceptual predictions. Specifically, functional impairment seems to be a correlate or indicator of the mentioned ongoing pathology [25-27]. Present findings suggest that individuals with relatively little functional impairment may still be in a good overall condition in which also the cognitive system is still able to compensate the ongoing pathology and maintain cognitive functioning (i.e., thereby showing a reduced cognitive decline). In contrast, highly vulnerable individuals with considerable functional impairment may already suffer from substantial losses that are indicative of the upcoming terminal decline of the whole system [47, 48]. After reaching this critical developmental stage, these individuals will not be able to compensate anymore the upcoming breakdown of the cognitive system (not even by large cognitive reserve). As detailed above, especially individuals with greater cognitive reserve accumulated will then show a much steeper cognitive decline because they will drop from a much higher initial level (than those with less cognitive reserve who have started to decline already relatively early) [4, 5].

One may argue that our measure of functional impairment (that based on difficulties in independently managing one’s activities of daily living), although being generic, yet could include a wide span of patients and some of them could be seriously compromised. To address this important issue, we controlled our analyses for chronic diseases participants suffered from and show that our findings are robust regarding such health-related confounders. Moreover, we demonstrate that already with a moderate functional impairment in some few activities (i.e., those individuals were still able to independently manage the majority of their other activities), the longitudinal relationship between greater cognitive reserve and steeper cognitive decline becomes significant. This suggests that this latter pattern may not be limited to individuals that are entirely dependent in all their activities of daily living. Future research might further scrutinize the role of additional factors including lifespan sociodemographic contexts, life events, personality, and age stereotypes that have been previously found to be important for cognitive aging [30, 33, 49, 50] and that may play a role in further shaping the pattern observed in the present study.

Our findings may appear somehow in contrast to recent cross-sectional studies in which cognitive functioning level among the more vulnerable individuals (i.e., those with poorer overall functional fitness status) seemed to more strongly depend on cognitive reserve [28, 29]. Yet, these latter studies differ in several ways from the present one: their design was only cross-sectional and considered rather momentary markers of cognitive reserve. Moreover, those studies focused on a relatively young sample of older adults (about 70 years on average), which may still have been able to largely compensate their beginning losses (i.e., in these studies, findings were explained by cross-domain compensation effects, such as cognitive reserve still allowing to overcome the detrimental cognitive aftereffects of poor functional fitness status). In contrast, in the present study, our sample was on average already around 80 years at the second assessment – an age at which usually many individuals already show substantial losses [31], and for which functional impairment, such as difficulties in independently managing one’s activities of daily living, may constitute a more pertinent issue [41, 42]. The present longitudinal study helps to advance our knowledge in this regard and suggests that individuals with greater cognitive reserve accumulated across the lifespan show a reduced cognitive decline if they still have relatively little functional impairment, while they will show a steeper decline (compared to individuals with less cognitive reserve) as soon as functional impairment becomes substantial.

“Published in Celebration of the 30th Anniversary of the inception of Dementia and Geriatric Cognitive Disorders 1990–2020”.

The authors are grateful to the Swiss National Science Foundation for its financial assistance. The authors also thank the participants of the VLV study, as well as all members of the LIVES project IP213 and LINK institute who contributed to the realization of the VLV study.

All participants gave their written informed consent for inclusion in this study before participating. The present study was conducted in accordance with the Declaration of Helsinki, and this study protocol had been approved by the Ethics Commission of the Faculty of Psychology and Social Sciences of the University of Geneva (project identification codes: CE_FPSE_14.10.2010, and CE_FPSE_05.04.2017).

The authors have no conflicts of interest to disclose.

This work was supported by the Swiss National Center of Competence in Research LIVES – overcoming vulnerability: life course perspectives, granted by the Swiss National Science Foundation (Grant No. 51NF40-185901). AI acknowledges support from the Swiss National Science Foundation (Grant No. 10001C_189407). E.R.G. and B.R.G. acknowledge support from LARSyS – Portuguese national funding agency for science, research, and technology (FCT) Pluriannual funding 2020–2023 (Reference: UIDB/50009/2020).

The funding sources had no role in the preparation of data or the manuscript. Moreover, the authors have not entered into an agreement with the funding organization that has limited their ability to complete the research as planned and publish the results. The authors have had full control of all the primary data.

A.I., E.R.G., and B.R.G. formulated the research question, analyzed the data, and wrote the manuscript. D.O. assisted in writing. M.O. and M.K. formulated the research question, conceptualized the study, supervised the data collection, and participated in writing.